top of page

From Experiment to Expression: Creating Music with Google's Lyria

Today, I embarked on an exciting journey of music creation using Google's Lyria. What started as a simple experiment turned into a rewarding process of crafting a track that truly reflects where I am in my life and creative path. After several iterations, I finally produced a piece I am proud to share. This post walks you through my experience, the creative decisions I made, and the story behind the music.


Eye-level view of a digital audio workstation screen showing a music track timeline and waveform
Creating music with Google's Lyria, showing the digital audio workstation interface

Starting with a Clear Vision


Before opening Lyria, I had a clear idea of the tempo and genre I wanted for the track. I aimed for an instrumental piece that blends ambient and electronic elements, something that could express the momentum I feel in my life right now. This clarity helped me craft precise prompts for Lyria, guiding the AI to generate music that matched my vision.


The inspiration came partly from a friend’s recent post about "building momentum." That phrase resonated deeply with me, capturing the energy and progress I’ve been experiencing. I wanted the music to reflect that feeling — a steady build-up, a sense of movement, and a hopeful tone.


The Iterative Process: From Rough Ideas to Final Track


Creating the track was not a one-step process. It took 3 to 4 iterations to get to a version I was satisfied with. Each iteration involved refining the prompts, adjusting the tempo, and tweaking the style to better capture the mood I wanted.


Here’s how the process unfolded:


  • First iteration: I started with a broad prompt describing the genre and tempo but left room for Lyria to explore. The result was interesting but lacked focus and emotional depth.

  • Second iteration: I refined the prompt to specify instrumental layers and a more defined rhythm. This version had better structure but still didn’t fully express the momentum I wanted.

  • Third iteration: I added more detail about the mood — emphasizing a gradual build and a hopeful, uplifting feel. This brought the track closer to my vision.

  • Final iteration: Small adjustments to tempo and instrumentation gave the track a polished, cohesive sound. This version felt complete and ready to share.


Each step taught me more about how to communicate with the AI and how to shape the music through careful prompting.


AI Music Production: Creating at the Speed of Thought


See how this final track evolved, from the first prompt until I got to a version which I find satisfactory. The Final Track


(Still In-Progress, will be uploading earlier iterations later...) Sharing the process feels important because it shows how experimentation and persistence lead to meaningful results. It also highlights how AI tools like Lyria can support creativity without replacing the human touch.


Next Steps: Making the Music Accessible


Once I finalize the track, I want to upload it to a music platform where others can listen and use it as background music for their own projects. I’m exploring options for creating an account on a site that supports sharing instrumental tracks freely.


This step will help the music reach a wider audience and provide a resource for content creators looking for original, mood-setting music. I’ll update this post with a link once the track is live.


Reflections on Using AI for Music Creation


Experimenting with Lyria has been a valuable experience. It showed me how AI can be a creative partner, offering new possibilities while still relying on human input for direction and emotion.


The key takeaway is that clear, thoughtful prompts make a big difference. The more specific I was about genre, tempo, and mood, the better the results. Also, being open to iteration allowed me to refine the music until it truly felt like an expression of my current journey.


This process has inspired me to continue exploring AI tools for creative projects and to share those experiences with others interested in music and technology.



 
 
 

Comments


bottom of page